Smooth and efficient integration of high-availability in a parallel single level store system

Abstract

A parallel single level store (psls) system integrates a shared virtual memory and a parallel file system representing an attractive support for long running parallel applications in a cluster. In this paper we present the smooth integration of a backward error recovery highavailability support into a psls system. Our highly-available psls system relies on a high degree of integration and re-usability between highavailability and standard supports. We focus on the parallel file system management at checkpointing and recovery time. A prototype has been implemented and we show some performance results.